How Did Polyphony Digital Pull Off Graphics and Dynamic weather on PS4

[05XR8]

Certainly would not bother me in all modes bar Custom Race.

 

[Shingo_civic]

They didn't say that anywhere.

Exactly. Kaz actually stated that it will affect the physics,it is not just a visual thing.

Some quotes from the Eurogamer interview:

That's all good news to me. The recent blog post said something about Gran Turismo having a different approach to the weather system and how it works. I wonder if you could talk through a bit more about how dynamic weather works in Gran Turismo 7 - does it impact track surface temperature, if there's a dynamic track, does it rubber in over the course of an event?

Kazunori Yamauchi: We don't have a system for simulating the amount of rubber that's laid on the track, but we do have a simulation of where the rain would create a level of water over the track. When a car passes over it, it will spray aside, and so the racing lines will dry quicker than the other parts of the track. That's something that is simulated and does affect the grip of the course accordingly as time goes on, and the track begins to dry in different places. The air temperature, the humidity, and then the road surface temperature will also be affected by these weather changes as well.

It seems it will be a very in depth weather system from what it seems

 

[05XR8]

Which means we finally get this confirmed from an old interview:

“Kazunori made one particularly interesting comment about surface traction. Polyphony’s capture teams do collect traction data from the tarmac surfaces of the tracks which they scan. However, variable surface traction is not currently modeled in Gran Turismo Sport. The fact they do have this information is noteworthy and suggests that it might be something we see in Gran Turismo in the future.”

In a thread about making old cars feel old, a discussion of tyres came up. I suggested since PD have this track surface data, we may see them do away with multipliers. Just as a thought, if PD have the data of say, Interlagos is more abrasive than Bathurst, we would know soft tyres wear quicker at Interlagos. The main factor would be track surface temp or ambient temp, however PD want to program it. Thus, we wouldn’t need to have 20x wear.

Fuel could factor in by quicker lap times. The faster you go, do to more grip, the more fuel is burned. Even when wear comes into play with spinning the wheels due to too much throttle. Again, however PD want to program it, the multiplier wouldn’t be needed.

here’s the interview https://www.gtplanet.net/forum/thre...ur-of-polyphony-digital.390761/#post-12922379

 

[JTB10000]

Eh, track surface data probably only means wet lines like Kaz mentioned in one of the recent interviews. I'm not expecting too much more complexity in the physics than GTS when they have to compromise for the PS4.

 

[ScottPuss20]

Eh, track surface data probably only means wet lines like Kaz mentioned in one of the recent interviews. I'm not expecting too much more complexity in the physics than GTS when they have to compromise for the PS4.

There's no reason why they couldn't alter the abrasion level for each track. I'm sure that it's possible for them to tweak it based on the surface data they've captured without breaking the PS4. Put it this way: the grip level and abrasion of each track is rated separately out of 10, and each number has a multiplier attached to it.
1 - 1x
2 - 1.1x
3 - 1.2x
4 - 1.3x
5 - 1.4x
6 - 1.5x
7 - 1.6x
8 - 1.7x
9 - 1.8x
10 - 1.9x
For grip levels, the higher the number the faster the track. For abrasion levels, the higher the number the coarser the surface. Important thing to note: grip level and abrasion are related so a track with high grip will tend to have higher abrasion. Example: Fuji Speedway has a grip level of 10 and an abrasion level of 8. The opposite is also true. Example: Willow Springs has a grip level of 4 and an abrasion level of 4. However this does not take temperature into account, which can skew these values. It is surface data only. High temps can speed up abrasion and decrease grip, whilst colder temperatures can reduce abrasion and increase grip. From a computational standpoint, this is probably the best option.

 

[05XR8]

Exactly.
Discussion in another thread brought that some players rely on the multiplier for testing. That’s fine, but knowing the abrasion level/rating and time of day, help just as well.
Nurburgring GP Noon during Summer and Noon during Winter, would yield Different results.

There’s also the dry line factor. In a 30 minute FIA race, do some players chance switching to RH/RM/RS or stay on IM/W? A smooth track like Bathurst could throw all kinds of strategies due to inclement conditions.
Or St Croix. Such a wide circuit, a dry line may take longer to form.

Other games do it. So, if PD have all this data to aid the realism, plus the grip meter, plus the changeable weather, it’s possible they’ve found a way to program some of these characteristics.

 

[ICEYOU3]

Wonder if the tire smoke will be improved.

 

[JTB10000]

Yeah, It's probably not as complex as I was thinking. Hopefully this is the game where they put all of the data into use. All of the pessimism surrounding this game has rubbed off on me and has me expecting the bare minimum from PD lol

 

[05XR8]

That may still be the case. Who knows? Until gtp et al. get the game before the rest of us do, only thinking about the possibilities.

 

[anthonyfa18]

After watching this i am still very scared how it will run on a base ps4, also for the replays i don't see a base ps4 hit 30fps

but i also feel ps5 will be not used to the max because of the game being cross platform

 

[Magog]

After watching this i am still very scared how it will run on a base ps4, also for the replays i don't see a base ps4 hit 30fps

but i also feel ps5 will be not used to the max because of the game being cross platform

Games like ACC are available on PC for the highest end rigs and yet don't take advantage of all that horsepower. We have reached an era in gaming where the financial resources and talent of the developers have far more to do with how good a game looks than the hardware they are running on. When it releases it will be one of the best looking racing games anywhere because they had the time money and talent to make it so.

 

[anthonyfa18]

So after yesterday show case of GT7, i really feel worried about players on ps4. i am happy i have a ps5
i feel the ps4 jaguar cpu will just melt🔥, even on ps4 Pro🧨. if some how this work i tip my hat to pd for pulling it off

 

[warp9engage]

Many times before, we've been impressed by Polyphony's prowess with dated hardware. I also wouldn't buy into the marketing hype for PS5, as they'll use language like "unparalleled realism and simulation", to make it sound like it's something completely out of this world new. Most of the time, it isn't that groundbreaking.

Horizon Forbidden West looks surprisingly good on PS4 Pro, at least. Polyphony, like Guerilla, is a first-party studio for Playstation. I do believe all first-party studios will be told to uphold an acceptable level of quality across all hardware.

 

[anthonyfa18]

A very Good Article to read
Lack of Gran Turismo 7 ps4 gameplay is an issue from racinggames.gg

The original PS4 is nine years old now. It goes without saying that the PS4 version won’t match the performance and visual fidelity of GT7 on PS5. The PS4 version likely has graphics and performance modes, running at 30 fps with graphics prioritised or 60 fps at a lower resolution. But Sony still hasn’t clarified this. To make GT7 run smoothly on PS4, compromises will have been made.

 

[Imari]

A very Good Article to read
Lack of Gran Turismo 7 ps4 gameplay is an issue from racinggames.gg

Not really. They're just guessing and they're doing it in a pretty aggressively controversial way, presumably to generate clicks. As you can see, it worked on you.

The reality is that GTS looks fine on PS4 and there's no reason to think that GT7 will be any different. FH5 runs on the X1, the one that has even worse hardware than the PS4, and it looks fine too. This idea that it'll be running at 30fps has no basis other than scaremongering - GT5 and GT6 were a long time ago and Polyphony has since demonstrated that they understand that a solid 60fps is important for gameplay.

Sony isn't showing PS4 footage because it doesn't look as good as PS5 footage. That doesn't mean it looks bad, it just means it's not as good for marketing purposes. PS4 is obviously going to look worse than PS5, but unless it's somehow significantly inferior to GTS or other similar games on similar hardware it's not a problem. There's no reason to believe that Polyphony's extremely selective day/night and weather systems will cause actual problems.

There's plenty of stuff to be concerned about for GT7 if you want, but this isn't really one of them. Maybe the PS4 release turns out to be a trash fire, but just because CDPR **** the bed isn't a reason to assume that Polyphony will.

 

[Famine]

A very Good Article to read ... from racinggames.gg

Does not compute.

 

[Magog]

Many times before, we've been impressed by Polyphony's prowess with dated hardware. I also wouldn't buy into the marketing hype for PS5, as they'll use language like "unparalleled realism and simulation", to make it sound like it's something completely out of this world new. Most of the time, it isn't that groundbreaking.

Horizon Forbidden West looks surprisingly good on PS4 Pro, at least. Polyphony, like Guerilla, is a first-party studio for Playstation. I do believe all first-party studios will be told to uphold an acceptable level of quality across all hardware.

Forbidden West looks "good" on PS4 and unbelievably jaw dropping on PS5. GT7 will likely be the same.

 

[MagpieRacer]

I suspect GT7 will look and perform fairly similar to GTS, and simply drop the res when the screen gets busy with weather and stuff like that.

 

[solidboss07]

Perhaps it was lost in translation and Kaz was describing how moisture laden air above the oceans would be drawn in overland by Playstation 4s' fans?

 

[Boxster]

For those worried about PS4 performance: you should look at this from a PD commercial perspective. If PS4 performance is bad, that's not good business for them. It would also be really bad PR for them. So I would be really surprised if this would turn out to be the case. Not impossible, but improbable.
Now what if PS4 performance isn't so bad. What do you do? You prefer to sell the PS5 version. So you really stress in a subtle way how good PS5 is and you show PS5 promotion material exclusively. I have been surprised how little PD makes of the PS5 benefits. Ray tracing the most important, and some other visual stuff.
So my guess is the PS4 will be fine as racing experience, but the graphics are a bit less impressive. I may be wrong here, but that's my guess.

 

[Veinz]

Part of me hopes it’s bad so this never happens again.

 

[toastet]

I have been surprised how little PD makes of the PS5 benefits. Ray tracing the most important, and some other visual stuff.
So my guess is the PS4 will be fine as racing experience, but the graphics are a bit less impressive. I may be wrong here, but that's my guess.

Raytracing only on replays on PS5. FH5 also looks good on Xbox One and GT Sport is still nice. I wouldn't worry too much.

 

[Scaff]

For those worried about PS4 performance: you should look at this from a PD commercial perspective. If PS4 performance is bad, that's not good business for them. It would also be really bad PR for them. So I would be really surprised if this would turn out to be the case. Not impossible, but improbable.

Laughs in GT5/6 Standard Cars and Tracks.

 

[TorqueHappenS08]

That complete time of day cycle with highly realistic dynamic volumetric clouds is seriously impressive. Not to mention all the new track side detail, foliage and dynamic grass as seen in the Tsukuba footage where the grass is blowing in the wind by itself.

It’s highly evident that PD did not hold anything back for the PS5 version, especially if you seen the 4K state of play Footage, absolutely beautiful.

I’m curious as to what the PS4 version looks/plays like.

 

[Magog]

That complete time of day cycle with highly realistic dynamic volumetric clouds is seriously impressive. Not to mention all the new track side detail, foliage and dynamic grass as seen in the Tsukuba footage where the grass is blowing in the wind by itself.

It’s highly evident that PD did not hold anything back for the PS5 version, especially if you seen the 4K state of play Footage, absolutely beautiful.

I’m curious as to what the PS4 version looks/plays like.

That's why I am giving them some slack for not having a dozen new tracks. They had to put that much effort into revamping the huge amount of tracks in Sport to get them up to that level of quality.

 

[Griffith500]

I've never seen clouds like that in a game before, mod or not.

I'd have to resort to real world time lapse radar to make it work in something like terragen even.

Incredible.


Needless to say PS4 won't get volumetric clouds, but they may still be able to generate and animate a cloud texture to overlay on the sky, similar to the raindrops on glass simulation in GT5.

 

[anthonyfa18]

 

[warp9engage]

I've never seen clouds like that in a game before, mod or not.

I'd have to resort to real world time lapse radar to make it work in something like terragen even.

Incredible.


Needless to say PS4 won't get volumetric clouds, but they may still be able to generate and animate a cloud texture to overlay on the sky, similar to the raindrops on glass simulation in GT5.

Well, I can say, with confidence, I know of at least one game that puts those clouds to shame. It will be making an appearance further below in this post.

From what I observe, it does seem GT7, at least, mimics volumetric clouds. I don't think anyone, besides Polyphony, knows exactly how complex, or simplified, this weather system actually is, or what trickery they might be using to make it look the way it looks, but there are a few things you can notice that suggests the system is toned down somewhat, possibly allowing it to work in older hardware, thus ensuring the feature parity that Yamauchi claims.

The first thing I noticed was the resolution of the clouds. I took a screenshot from the GT7 State of Play, and then did the same with a GT Sport video, to attempt to show this. Both at 4K. I also tried to get the same track and the same position. Unfortunately, the same exact weather conditions were not available for both. Still, there are cloud formations you can look at and compare. What we have are clouds with a very similar texture resolution, i.e. not exactly high.

GT7 clouds at Spa

GT Sport clouds at Spa

The second thing I noticed comes from the various time-lapses, shown during the GT7 State of Play, in the "Time & Weather Simulation" section. If you observe the clouds during the time-lapse that starts at the 13:38 mark, and on another time-lapse at the 13:57 mark, you'll notice they change shape in a rather clunky manner, despite the footage being sped up. This suggests that, at least, part of the weather simulation that leads to these clouds, is running at a low refresh-rate. Compare this to how Microsoft Flight Simulator's clouds behave and change their shapes, under a time-lapse, and you can see what a truly next-gen cloud and weather system can look like.

So, in short, I think Polyphony is building this weather system with lower resolution assets and lower simulation refresh-rates, in order for it to be within the reach of the PS4's processor, thus allowing feature parity between the platforms.

 

[Vspectra]

Don't even care how the PS4 versions will perform, just be glad dynamic time and weather is back.

 

[Scaff]

Well, I can say, with confidence, I know of at least one game that puts those clouds to shame. It will be making an appearance further below in this post.

From what I observe, it does seem GT7, at least, mimics volumetric clouds. I don't think anyone, besides Polyphony, knows exactly how complex, or simplified, this weather system actually is, or what trickery they might be using to make it look the way it looks, but there are a few things you can notice that suggests the system is toned down somewhat, possibly allowing it to work in older hardware, thus ensuring the feature parity that Yamauchi claims.

The first thing I noticed was the resolution of the clouds. I took a screenshot from the GT7 State of Play, and then did the same with a GT Sport video, to attempt to show this. Both at 4K. I also tried to get the same track and the same position. Unfortunately, the same exact weather conditions were not available for both. Still, there are cloud formations you can look at and compare. What we have are clouds with a very similar texture resolution, i.e. not exactly high.

GT7 clouds at Spa
View attachment 1113054

GT Sport clouds at Spa
View attachment 1113055

The second thing I noticed comes from the various time-lapses, shown during the GT7 State of Play, in the "Time & Weather Simulation" section. If you observe the clouds during the time-lapse that starts at the 13:38 mark, and on another time-lapse at the 13:57 mark, you'll notice they change shape in a rather clunky manner, despite the footage being sped up. This suggests that, at least, part of the weather simulation that leads to these clouds, is running at a low refresh-rate. Compare this to how Microsoft Flight Simulator's clouds behave and change their shapes, under a time-lapse, and you can see what a truly next-gen cloud and weather system can look like.

So, in short, I think Polyphony is building this weather system with lower resolution assets and lower simulation refresh-rates, in order for it to be within the reach of the PS4's processor, thus allowing feature parity between the platforms.

For comparison these are ACs modded volumetric clouds, a mod created by one person.